A working prototype of grid system for storage and intelligent processing of large amounts of ECG signals in a wide scale of Ukraine's population, implementation of which in Ukraine will carry out important population researches and to identify the features of the evolution of cardiac diseases that are specific for different regions of Ukraine was developed.
For the first time in Ukraine through telemedicine and grid technologies there was conducted a pilot population-based study of electrocardiograms and there were identified risk factors for the Poltava region of Ukraine.
The launching technology on the grid-resources “Medgrid” virtual organization of virtual machines with the ability to perform interactive processing programs of cardio signals was established. This result can be regarded as a combination of “grid-computing” paradigm with “cloud computing” paradigm. The interaction of “Medgrid” services project with Ministry of health telemedicine portal (esemi.org) was demonstrated and infrastructure interaction with interacting medical information system of the hospital for NAS scientists was started as well.
The method, algorithm and program approximation of plane and spatial experimental curves, in particular, vectorcardiograms in order of their analytical presentation using preset parametric splines were developed.
Grid applications and grid working prototype systems for expansion of digital diagnostic data range, in particular, spirogram based on grid cluster of IMMSP NAS of Ukraine with computer resources virtualization for organization of interactive work with digital electrocardiograms.
A new additional Web Service “Telemed” providing automation of depersonalized reception and storage of digital diagnostic data was created. To ensure remote signal processing by doctor-consultant it was developed a dynamic service run of virtual machines on a grid platform (rainbow technology) enabled links to diagnostic data grid storage.
For the first time for applied problems of medical informatics the algorithm and software implementation of approximation curves parametric spline curves in three-dimensional space coordinates were developed. The software implementation was developed and the effectiveness of the algorithms identifies of a person according to his electrocardiogram in 3 orthogonal and 6-lead base was studied experimentaly.
The algorithms and programs for identification (authentication) of human for his (her) electrocardiogram were developed. In particular, it was developed the new method of calculating the trajectory of averaging the QRS-complex 3D-space coordinates and a comparative study of sensitivity and specificity of the original algorithm for identifying human electrocardiogram in three orthogonal leads and the first three standard leads. An average sensitivity of original algorithm was about 97% and 95% respectively.
Based on the known neurophysiological researches on the functioning of the visual system it is designed and modeled a new method of software segmentation hurt distorted complex signals, which involves the use of variable resolutions and does not use a priori information concerning the signal and noise distorting it. The experiments using this method for segmenting complex signals, including cardiograms and spirogram were done.
There were carried out the experiments to determine the effectiveness of biometric identification of human electrocardiogram and Modification of the original copyright identification algorithm to conditions of grid technologies Rainbow, which allows dynamic service run virtual machines on a grid platform.
On the base of a new biometric method for processing visual information based on the use of variable resolution, a new method and algorithm for the segmentation of distorted disturbances of experimental data, which are differ from the known ones (they do not use a priori information about the properties of interferences) have been developed. This development is software implemented and tested on model signals and real signals of single-channel cardiograms. The program implementation of the algorithm of biometric identification of a person is developed on its multichannel electrocardiogram. The dynamic library is developed, which allows implementing the algorithm of biometric identification in the process of electrocardiogram processing in the "Medgride" project.
On the base of cloud architecture using a pilot web portal of a family doctor was developed and launched to evaluate the level of health of his patients based on the rapid analysis of electrocardiograms. Information technologies developed in the Medgride project can be considered as an active prototype of modern infrastructure for the safe storage of any volume of depersonalized medical diagnostic information in the scale of the population of Ukraine.
The elements of information technology for identifying a person to his or her electrocardiogram, the segmentation of electrocardiograms on cardiocycles, used to automate the functional tasks of telemedicine, mobile medicine and home telemedicine, are based on new diagnostic devices, in particular mobile cardiographs with reduced number of leads. Based on the use of cloud architecture, a pilot web portal of a family doctor was developed and continued to be used to assess the health of his patients based on the rapid analysis of an electrocardiogram with a reduced number of leads.
Developed, software implemented and experimentally tested algorithm for human authentication on its single-channel electrocardiogram and algorithm for segmentation of the cardio signal at variable resolution in real time, which does not require additional settings for interference parameters, which allows the use of these algorithms in other remote applications.
The minimum hardware infrastructure for the private cloud of IPMMS of NAS of Ukraine is created. Private cloud services have been deployed and configured, which will be used in IPMMS projects of the NAS of Ukraine in the most important areas of research, including for the needs of “the Medgrid” project.
Automated information technology of respiratory control «Harmony», which increased the effectiveness of psychological rehabilitation of people with high levels of anxiety, was developed, tested and implemented in the center of psychological rehabilitation. The technology can already be used in psychological rehabilitation centers, psychological rehabilitation rooms, specialized sanatoriums, etc. Further research of the technology allows us to expect new results in the field of functional diagnostics and monitoring of the functional state of man.
A cloud service of human biometric authentication according to its multichannel and single-channel electrocardiogram as a component of a complex of interacting cloud information technologies for the accumulation and intelligent processing of ECG examinations was developed and experimentally tested. Widespread introduction of such technologies in Ukraine will allow in the medium term to move to the so-called personalized telemedicine.
A method of a three-dimensional representation of cardiocycles was created using the method of approximating experimental curves with canonical splines, developed at the IMMSP of the National Academy of Sciences of Ukraine. It made it possible to make effective algorithms for the identification and authentication of people based on their ECG. This method of person authentication is gaining popularity due to the fast spread of telemedicine technologies in the paradigm of personalized home and mobile medicine.
Cloud services of the first stage of the Medgrid monitoring project were developed. They ensure an assessment of a person's condition according to the parameters of coherence of heart rate variability and biometric authentication of the patient based on his/her electrocardiogram.
An original algorithm for automatic marking of breath hold phases in hypoxic samples of Stange and Genchi using a single-channel electrocardiogram was developed, which will be used in prototypes of information technologies and cloud services of the Medgrid project.
The architecture was improved and the corresponding services of the information system in the cloud environment were developed to solve the problems of both practical health care and support to scientific decentralized research, which will allow to accumulate data of users' cardiograms for a long time, perform their processing and analysis, provide access to data arrays in accordance with the principles of FAIR (findable, accessible, interoperable, reusable), as well as make the correction of psycho-emotional state of a person.
An algorithmic basis for processing the cyclic signal of a single-channel electrocardiogram in terms of the automatic selection of respiration signals was developed, and an experimental evaluation of the effectiveness of algorithms for assessing the psycho-emotional state of a person during psychological rehabilitation was conducted. The proposed segmentation algorithm will be implemented as a cloud service for the information system supporting long-term monitoring studies within the Medgrid project.
